Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Isometric Propagation Network for Generalized Zero-shot Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add visual prototypes to the semantic-prototype classifier (ablation) | IPN · AWA2 | unseen-class accuracy: 61.1% to 67.5% | Helped |
| use only visual prototypes instead of semantic prototypes (ablation) | IPN · AWA2 | unseen-class accuracy | Hurt |
| apply propagation only in the visual space (ablation) | IPN · AWA2 | unseen-class accuracy: 2~3% | Helped |
| apply propagation only in the semantic space (ablation) | IPN · AWA2 | unseen-class accuracy: 2~3% | Helped |
| remove propagation from both spaces (ablation) | IPN · AWA2 | unseen-class accuracy | Hurt |
| add dual propagation in visual and semantic spaces (ablation) | IPN · AWA2 | H | Helped |
| replace separate attention modules and consistency regularization with one shared visual attention module (ablation) | IPN · AWA2 | seen-class overfitting | Hurt |
| replace separate attention modules and consistency regularization with one shared semantic attention module (ablation) | IPN · AWA2 | seen-class overfitting | Hurt |
| remove consistency loss while keeping separate attention modules (ablation) | IPN · AWA2 | generalized ZSL performance | No clear effect |
| vary the propagation steps from the selected value of 2 (ablation) | IPN · AWA2 | generalized ZSL performance | Mixed |
| vary the temperature γ (ablation) | IPN · AWA2 | accuracy | No clear effect |
Closest papers by meaning
- Attribute Propagation Network for Graph Zero-Shot Learning (AAAI 2020 · similarity 0.93)
- Graph Knows Unknowns: Reformulate Zero-Shot Learning as Sample-Level Graph Recognition (AAAI 2023 · similarity 0.89)
- Adaptive and Generative Zero-Shot Learning (ICLR 2021 · similarity 0.89)
- Improving Generalized Zero-Shot Learning by Exploring the Diverse Semantics from External Class Names (CVPR 2024 · similarity 0.88)
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.87)
- Data-Free Generalized Zero-Shot Learning (AAAI 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.